Apple N1 Wi-Fi Chip Improves On Older Broadcom Chips In Every Way (arstechnica.com) 34
An anonymous reader quotes a report from Ars Technica: This year's newest iPhones included one momentous change that marked a new phase in the evolution of Apple Silicon: the Apple N1, Apple's first in-house chip made to handle local wireless connections. The N1 supports Wi-Fi 7, Bluetooth 6, and the Thread smart home communication protocol, and it replaces the third-party wireless chips (mostly made by Broadcom) that Apple used in older iPhones. Apple claimed that the N1 would enable more reliable connectivity for local communication features like AirPlay and AirDrop but didn't say anything about how users could expect it to perform. But Ookla, the folks behind the SpeedTest app and website, have analyzed about five weeks' worth of users' testing data to get an idea of how the iPhone 17 lineup stacks up to the iPhone 16, as well as Android phones with Wi-Fi chips from Qualcomm, MediaTek, and others.
While the N1 isn't at the top of the charts, Ookla says Apple's Wi-Fi chip "delivered higher download and upload speeds on Wi-Fi compared to the iPhone 16 across every studied percentile and virtually every region." The median download speed for the iPhone 17 series was 329.56Mbps, compared to 236.46Mbps for the iPhone 16; the upload speed also jumped from 73.68Mbps to 103.26Mbps. Ookla noted that the N1's best performance seemed to improve scores most of all in the bottom 10th percentile of performance tests, "implying Apple's custom silicon lifts the floor more than the ceiling." The iPhone 17 also didn't top Ookla's global performance charts -- Ookla found that the Pixel 10 Pro series slightly edges out the iPhone 17 in download speed, while a Xiaomi 15T Pro with MediaTek Wi-Fi silicon featured better upload speeds.
While the N1 isn't at the top of the charts, Ookla says Apple's Wi-Fi chip "delivered higher download and upload speeds on Wi-Fi compared to the iPhone 16 across every studied percentile and virtually every region." The median download speed for the iPhone 17 series was 329.56Mbps, compared to 236.46Mbps for the iPhone 16; the upload speed also jumped from 73.68Mbps to 103.26Mbps. Ookla noted that the N1's best performance seemed to improve scores most of all in the bottom 10th percentile of performance tests, "implying Apple's custom silicon lifts the floor more than the ceiling." The iPhone 17 also didn't top Ookla's global performance charts -- Ookla found that the Pixel 10 Pro series slightly edges out the iPhone 17 in download speed, while a Xiaomi 15T Pro with MediaTek Wi-Fi silicon featured better upload speeds.
Good to see (Score:5, Insightful)
Any good competition for Broadcom is a plus for users. I expect Apple had to wade through a thicket of Broadcom patents in order to make their chip, not a trivial accomplishment.
Re:Good to see (Score:5, Insightful)
The only people who see any benefit are Apple customers, but If place this into the category of nice to have, but not the kind of thing I'd expect to notice. No one is particularly threatened by Apple's chip being better because Apple doesn't sell chips, they sell iDevices.
Re: (Score:3)
Re: (Score:3)
WiFi patents are FRAND out of necessity, and I believe they're available in patent pool form, so all Apple has to do is take out a license with the patent pool to acquire a license to every patent they need for WiFi.
Re: (Score:3)
There's a big difference between "necessary for WiFi" and "necessary to do WiFi well". Patents often cover the later rather than the former.
Re: (Score:3, Interesting)
Indeed, there are a great many trade secrets in the RF business. However, I expect all this to level out in the coming years. Physics provides a limited spectrum, and the unlicensed and licensed sides in this are already squabbling over what spectrum there is, because all the useful bands (<=6-7GHz) are now allocated, somehow, to one side or the other.
The Wi-Fi people understand this: Wi-Fi 7 already covers all the unlicensed spectrum that isn't still being squabbled over, and even some that is. Wi-
Re:Good to see (Score:4, Interesting)
Apple's usual problem with FRAND patents is that it's own patent portfolio is largely worthless. Nobody is going to exchange a valuable WiFi patent for a design patent on rounded corners. Apple could just pay the licencing fees, but doesn't like to.
It's not just Broadcom they have to licence from either. A lot of the WiFi 6 and 7 stuff was invented by Huawei. Apple's cellular modems require paying Huawei too.
not a shock (Score:3, Interesting)
New radio older than old radio? Wow.
What I find actually surprising is not in the headline but is in the summary: Mediatek is superior.
Re: (Score:3)
New radio older than old radio? Wow.
What I find actually surprising is not in the headline but is in the summary: Mediatek is superior.
Heh. i assume you mean "faster". Still it's an accomplishment to be competitive for a first foray into the market. (of wifi / BT chips).
That said.... i'm pretty skeptical of the methodology here. reading the ookla article this really IS only based on end-to-end speed tests. There's splits by market (continent) and some talk of Wifi7 vs not. But there's no talk of any other controls like for type of internet access, type of access points , etc... presumably because none of that info is avail
Re: (Score:1)
Yeah, that was a big goof, thanks for understanding.
Apple is capable of hiring talented people and creating a useful product. They just don't seem to be capable of being user-friendly in the ways that matter to me. TBH they were never great at it, and MUGs did the heavy lifting in the customer relations department for them for free. Anyway I'm totally capable of believing their performance claims, to a reasonable point, especially when the results aren't putting them first.
I wish they were friendlier, becau
Re: (Score:2)
Yeah, that was a big goof, thanks for understanding.
Apple is capable of hiring talented people and creating a useful product. They just don't seem to be capable of being user-friendly in the ways that matter to me. TBH they were never great at it, and MUGs did the heavy lifting in the customer relations department for them for free. Anyway I'm totally capable of believing their performance claims, to a reasonable point, especially when the results aren't putting them first.
I wish they were friendlier, because their hardware is reasonably impressive. I'm also just not in their target demographic apparently because I'd rather have a slightly thicker device with better cooling and battery capacity.
And who, pray tell, is CONSISTENTLY better at it?
Re: (Score:2)
New radio older than old radio? Wow.
What I find actually surprising is not in the headline but is in the summary: Mediatek is superior.
Heh. i assume you mean "faster". Still it's an accomplishment to be competitive for a first foray into the market. (of wifi / BT chips).
That said.... i'm pretty skeptical of the methodology here. reading the ookla article this really IS only based on end-to-end speed tests. There's splits by market (continent) and some talk of Wifi7 vs not. But there's no talk of any other controls like for type of internet access, type of access points , etc... presumably because none of that info is available.
They made no statements afaict as to why one would believe iPhone 16 and iPhone17 users are functionally similar in those regards. It seems at least reasonable to me to suspect that the people who sprung for the new $1000 phones in their first month would also be more likely to have faster internet service and better routers...
I think one thing being overlooked is the Thread Integration.
Apple is about to do a big Home Automation push, and Thread/Matter MESH capability in more devices is a part of that.
Re: (Score:2)
Considering that Apple has had several releases with worse radios than what they shipped immediately previous to those releases, this is kind of news.
Had they not fucked up spectacularly in the past, you would definitely be right.
New TV improves old TV in every way (Score:2)
No shit, Sherlock!
Re: (Score:2)
With Apple, it's not necessarily a guarantee that a new version will be better in every way than the old version.'
They've shipped several "new" versions with shittier radios than previous versions over the lifetime of the iPhone alone.
who has these DL speeds? (Score:2)
Not me. I already pay enough just to get 50 mbps down.
Re: who has these DL speeds? (Score:2)
I pay 33 quid ($43) per month for unlimited data and uncapped speeds with roaming to 83 countries and Netflix Premium as a bonus. I get hundred of mbps too.
Re: who has these DL speeds? (Score:2)
*hundreds
Thatâ(TM)s as-in nearly 300mbps at home where the 5G signal is weak
Re: (Score:1)
Satellite (Score:2)
In the future, phone to satellite as a backup will be important. I hope they are working on making that work properly. It sucks that there are so many areas that have zero service. We've been paying the universal service fee tax for decades now and there's nothing to show for it. I mean, even service in cities still have dead zones. This is 20 years after those "can you hear me now" Verizon ads when they claimed to be testing reception. Phone to satellite would in principle be a lot cheaper than terrestrial
Re: (Score:2)
That's already available on my 16Pro. Not sure whose chip they use, or its performance level, though.
Integrated into CPU? (Score:2)
Is Apple's end objective here to integrate the modem into the CPU so that there's one less chip in the device? That should save significant cost.
Re: (Score:2)
The biggest reason they don't do that is the CPU generates a large amount of EMI noise, the modem would have to deal with the crosstalk from parasitic coupling.
Re: (Score:2)
Intel already does this integration (within the limits you described): https://en.wikipedia.org/wiki/... [wikipedia.org]
It's a real PITA to pick an intel wireless module that will work with non-intel CPUs.
Re: (Score:2)
The biggest reason they don't do that is the CPU generates a large amount of EMI noise, the modem would have to deal with the crosstalk from parasitic coupling.
Exactly!
Not to mention that a failure in that one subsystem would ruin the whole Die.
So not worth it.
Re: (Score:1)
Is Apple's end objective here to integrate the modem into the CPU so that there's one less chip in the device? That should save significant cost.
That's the end objective of everyone, everywhere, at all times. Never bet against integration.
Re: (Score:2)
Probably not. My knowledge is probably out of date by now but it used to be the case that the top end processes not only would be wasted on analogue radio stuff, but they'd also need to be adapted to it a well. You want different features.
Re: (Score:2)
No, their 'end objective' here is to not have to pay Broadcom for a chip that Apple can make itself, while also having more control over what the chip does and doesn't do.
Same reason they started making their own CPUs rather than using PowerPC or Intel.
Re: (Score:2)
No, their 'end objective' here is to not have to pay Broadcom for a chip that Apple can make itself, while also having more control over what the chip does and doesn't do.
Same reason they started making their own CPUs rather than using PowerPC or Intel.
DingDingDing!!! We have a Winner!